Among their findings: results that blended search results - results that feature images or offer links to drill down on news or images - are clicked at twice the rate of conventional 'in-line' results. While companies don't often have multimedia results, they do have the opportunity to perform 'call-outs' to other sources of data in the company such as names and phone numbers. In addition, facets for guided or parametric search are similar to the 'news', 'images', and other 'invisible tabs' links on Google; so we can imagine that dynamic facets enjoy the same high rate of click-through. An unexpected find: fewer internet searchers bother with results past the first few result pages. A full 49% never explore past the first page, and the trend is growing. It's a study worth reading
NIE is happy to announce the new www.searchcomponentsonline.com, where you can browse for open source, low cost, and commercial components related to enterprise search. We're especially excited about this because, up until now, there has not been a central reference for all sorts of tools for just about every search out there. You'll find document filters, text processing tools, conversion utilities, extraction tools, and more. And we'll be adding more all the time.
If you have components and tools that you find useful, let us know, or leave a comment on www.searchcomponentsonline.com.
If you're the 'Dr Search' for your department or company, you need to get your very own Dr Search gear. Right now you can pick up a mouse pad or one of two sizes of coffee mug; and our Doctor is working on finding more cool stuff so everyone in your office knows that you are the Dr Search of your company. Check it out!
TechCrunch has a video if a cool integration of Web 2.0 and Search 2.0 that looks like a more useful extension of Wiki KeyMatch new to Google Appliance Version 5.
Basically users can vote for high quality results, and when such a result shows up in a search, there is a small green graphic adjacent to the result. Clicking on the result lets you add your own comment to the result, and provides a way to view what other people thought about the result. As long as users are honest with their evaluations, just having a small green tab by the result shows that at least one person thought the result was good.. In the corporation, this looks like a much better solution than existing Wiki KeyMatch, which just places users' comments in-line where 'best bets' or promoted content usually goes.
The TechCrunch article is called "The Future of Search?" Maybe a future.
A reminder that the Enterprise Search Summit West in San Jose, California, is early this year. Normally in early November, this year's conference and the concurrent KM World show will be held September 23-24 at the San Jose McEnery Convention Center. Maybe they decided that the election, scheduled for November 4, would be too distracting.
NIE will be exhibiting, and will be speaking on both Selecting a Search Engine and on a completely new topic, Search and the Virtual Machine. The Advance Program is available online.
Register now!
The Harvard Business Review has published research by Anita Elberse that casts doubts on the Chris Ancderson's Long Tail Theory.
Chris, who collaborated with Anita on his research, suggests that at least part of the difference is how one defines 'head' and 'tail' - thank goodness not on the meaning of 'is'. The article on the HBR has links to blogs by both authors, and it's worth a read.
Regardless of how their research proceeds, we are still firm believers in the 'long tail of search', and in how important it is to get your ' top 100' query terms working really well.